Urban Policy in Post-Apartheid South Africa: Context, Evolution and Future Directions

Abstract

Against a backdrop of definitions and conceptual clarifications of the term urban policy, including its expression in the developing world—and Africa, in particular—this paper reviews the trajectory of urban policy in South Africa post-1994 and comments on future directions and plausible scenarios. In a highly specific context-dependent analysis, the paper argues that, in the first 10 years of democracy, we have seen the creation of democratic, integrated development local government, mass delivery of housing and services, a finely crafted array of capital and operating subsidies for delivery to lowincome households, and a number of programmes intended to enhance the capacity of local government to undertake delivery. All of these have been centrally driven. The counterpoint to the national perspective and frameworks, the paper argues—and in evidence over the past 5 years, in particular—is the increasingly robust role and influence of cities in setting the urban agenda and, in effect, leading urban policy. The paper concludes by examining the many points that contribute to this view.

The Nature of Urban Policy

The development of most urban areas is influenced, to some degree, by the processes of urban policy and urban planning. Urban policy and planning are generally concerned with the management of urban areas. They are state activities that seek to influence the distribution and operation of investment and consumption processes in cities for the ‘common good’. However, it is important to recognise that urban policy is not confined to activity at the urban scale. National and international economic and social policies are as much urban policy if defined by their urban impacts, as is land use planning or urban redevelopment. In effect, urban policy is often made under another name. Urban policy and planning are thus dynamic activities whose formulation and interpretation are a continuing process. Measures introduced cause changes that may resolve some problems, but create others for which further policy and planning are required. Furthermore, only rarely is there a simple optimum solution to an urban problem. More usually, a range of policy and planning options exist from which an informed choice must be made.

Urban policy is also the product of the power relation between the different interest groups that constitute a particular society. Foremost among these agents are government (both local and national) and capital in its various fractions. Capital and government pursue specific goals that may be either complementary or contradictory. These political and economic imperatives have a direct influence on the nature of urban policy. Urban policy is also conditioned by external forces operating within the global system as well as by locality-specific factors and agents. It would also be fair to conclude that urban policy making is the product of a continuous interaction of intellectual process and institutional response, a process driven by successive sets of powerful and relatively consistent value judgements which have a profound influence on how urban problems are defined, and on the policies derived to deal with them.

Urban Policy in the Developing World: Influential Waves of International Research

Under the sponsorship of the World Bank who, as Rogerson (1989) writes, recognised that urbanisation problems in the developing world had reached proportions warranting serious consideration in overall national development policies and that the formulation of coherent national urban policies was undeveloped in most African, Asian and Latin American countries, a first wave of research in the early 1980s argued that national urban policies should be conceived as a subgroup of national spatial strategies primarily directed at urban rather than rural settlements. The objectives of national urban policy were viewed as multidimensional in character even if a few explicit goals might be stressed. Furthermore, due to the geographical concentration of investments (in urban areas) and the high costs of urban infrastructure, a set of national urban policy goals were often framed in terms such as slowing the growth of the primate city, strengthening intermediate cities or minimising urban–rural migration. The tendency to specify explicit spatial goals as the prime objective of national urban policy was severely criticised. Rather, in seeking objectives for national urban policy, it was essential to acknowledge that such goals could not be “ends in themselves but are the means to achieve the general aims of society” (Richardson 1983, cited in Rogerson 1989). Therefore, spatial objectives, for example slowing down the growth of the primate city or minimising rural–urban migration, were seen as only having merit as sub-goals in achieving higher level societal goals. This view therefore posited that the core objectives for national urban policy were the same as those of general national economic and social policies and included inter alia the promotion of economic growth and efficiency, improving equity and reducing poverty, satisfying basic human needs for all and preserving environmental quality. As Rogerson (1989) points out further, a number of influential studies at the time agreed that it would not be advisable to set goals for national urban policy that could be measured solely in quantitative terms, such as limits to population growth or targets for individual cities and regions. The overall policy goal would be to attain an optimal settlement pattern that maximises real income for people regardless of where they reside. In terms of the actual form of national policy, this body of work emphasised the inherent danger of assuming that a general strategy would be applicable to all developing economies. Instead, country uniqueness was emphasised and a typology of urban development strategies subsequently emanated.

In the late 1980s, a second wave of research began to question the value of direct location instruments in national urban policy, strongly arguing that indirect mechanisms may be more appropriate, and de-emphasising spatial considerations in favour of a ‘space-blind’ policy which becomes a key component of overall economic and development planning. The observation is made by a number of scholars that little evidence exists to support the view that experiments with national urban policy conducted from the mid-1970s to the mid-1980s (including those embracing the framework described as the first wave of research) have been successful. The most common explanation for the poor performance of national urban policies across the developing world is seen to be the separation of national urban policy from the national economic planning process and from the overall context of macroeconomic and sectoral policies. In other words, in most developing economies, national urban policy was often implemented in abstraction from overall development strategies, with very little synergy between the two processes. National urban policy was generally treated as a narrow sectoral responsibility for which a national line ministry had overall responsibility. There was consequently little integration between urban development policy and development planning in general. The urban system was moulded independently of other policies, resulting in a failure to recognise the powerful spatial impress exerted by macroeconomic forces on the pace of urbanisation and the form and functioning of the urban system. In sum, independently conceived spatial policy, for example those still fixated with the ‘primacy’ problem, still had strong resonance.

The African Experience: A Cursory Overview

The experience of African countries in formulating and implementing urban policy over the last two and a half decades has been haphazard. Urban growth has taken place exceedingly rapidly, with all evidence suggesting that urban populations will continue to grow much faster than rural populations even if the urban bias in development strategies were reversed. Whilst the rate of the continent’s growth may be lower than other regions—due primarily to Africa’s marginalisation in terms of trade, investment and infrastructure development (Halfani 1996)—and theories abound about a slowdown in the rate of growth of some of the largest cities and of polarisation reversal or spatial deconcentration into polycentric metropolitan forms (Pacione 2005), the fact of the matter is that Africa is becoming increasingly urbanised. What is therefore needed, argues Hope (2001), is the development and implementation of appropriate national urbanisation strategies, influenced from country to country by the economic, social, political and cultural characteristics that exists within each or, as the point has been made above, a set of strategies that takes into account the uniqueness and distinctiveness of the nation states, and indeed cities, in question.

Now whilst urban policy alternatives (see above section) were available over the last two decades with which urban policy makers in the developing world could experiment with, the urban situation in Africa has made it exceedingly difficult to implement a uniform set of policy precepts so that at any given time a range of interventions seemed to coexist, from spatially explicit urban policies in some parts of the continent to more ‘space blind’ initiatives elsewhere to yet a combination of the two in others. Perhaps, the exception has been Kenya where national urban policy, instead of being treated as a self-contained area of policy making (explicit spatial policy intervention to restructure the national urban system), has more recently rejected the premise that the urban system can be thought of independent of other policies and has recognised the formidable spatial impress that macroeconomic forces exert on the pace of urbanisation and the structure and operational efficiency of the Kenyan urban system. The point made by Evans (1989) here is that national urban policy in Kenya is no longer concentrated on the form and pattern of the urban system in the context of a postulated end-state; rather, the urban settlement system is viewed largely as an outgrowth of the macroeconomic policy context.

The unique and differing situation among African cities which, as pointed out above, has historically made it difficult to implement a set of policy principles in any sustainable manner results from a range of spatial and demographic factors. Three are worth mentioning here. Firstly, definitions of urban and rural vary widely across Africa. Many African countries use a population size of 2,000 as the cutoff between rural and urban settlements. However, the figure varies from 100 in Uganda to 20,000 in Nigeria and Mauritius. Almost half the countries in Africa use a numerical definition to indicate the areas that qualify as urban (UNCHS Habitat 2001). The second has to do with the different patterns of urbanisation in West and East Africa. In many West African countries, there are few secondary cities, resulting in the population being concentrated in one or a few large cities. Population growth in East Africa is more evenly distributed over secondary and tertiary cities. But there, also, primary cities are going through a period of rapid growth. The patterns of urbanisation in North and Southern (especially South) Africa have also been in stark contrast (Pillay 2004a, b). Thirdly, whilst an older body of literature has often described a general model of the African city based on the existence of an indigenous core and the distribution of different ethnic groups according to density gradients which assigned low-density land use to the administrative and residential requirements of the colonial elites and high density to indigenous populations, the situation in most African cities at present is hardly analogous. Post-colonial transformations of African cities, characterised by a greater mixing of economic and residential land uses, has resulted in a variety of forms (Pacione 2005), underscoring the point made above that any set of urban development strategies has got to factor in the unique and distinctive spatial morphology inherent in most African cities, as well as a parallel set of complex institutional and political arrangements.

South Africa

An Overview—Pre-1994

Urban policy pre-1994 in South Africa was based predominantly on the dictates of apartheid spatial planning, with the precise form of the South African city being codified by the 1950 Group Areas Act and the notion of segregated urban space. It is little wonder, therefore, that current urban policy is founded on an intention to reintegrate cities and move towards more compact urban forms. As Todes (2000) has observed, visions akin to the urbanist ideals of Jane Jacobs (1961) offering opportunities for higher density living, proximity between home and work, land use mix and social integration are prevalent. Alternative discourse of city development emerged as planners, and urban scholars mounted a huge critique of urban apartheid. Less divided urban forms were proposed in an integrated city and coordinated development framework. Negotiated during the inclusive forum processes that characterised South Africa’s transitional period (Pillay 2004a, b) and endorsed, as Todes (2000) goes on to reveal, by neoliberal technocratic bodies such as the Urban Foundation (1990) and World Bank (1991), these ideas soon became the dominant discourse and were effectively embodies in legislation by the new government. As Pieterse (2004) has remarked, urban integration was seen as providing the ultimate panacea to the many intractable problems that mark South Africa’s cities, emerging strongly in the ‘recently’ published Draft White Paper on Urbanisation and before that at the core of government’s 1997 Urban Development Framework. Harrison et al. (2003) note cautiously, however, that there is real and growing concern that government’s ‘neoliberal’ (market-orientated) turn may be exacerbating social and class divides and may be prioritising South Africa’s standing in the global economy at the expense of its poorest citizens. Robinson (1998) has argued too that such a technocratic discourse could become another oppressive form of urban ordering—a physicalist meta-narrative imposing a single moral view of the good city, a sentiment strongly echoed by Bond (2003) who has argued that the core characteristics of post-apartheid urban policy resulted, through unintended effects, in an equally oppressive structured process that can be termed class apartheid. As a result, he goes on to elaborate a variety of specific problems associated with apartheid-era urban underdevelopment continued—and were in may cases exacerbated—during the late 1990s and into the twenty-first century. In sum then, a growing body of literature has begun to argue that compaction-integration appears to offer little to the urban poor, especially those on the periphery of the city, and, in the context of a larger emphasis on development local government and more recently a powerful discourse on cities and global competitiveness, new urban policy formulations are seen to be worthy of investigation.

This is not to gainsay the importance of compaction-integration strategies in opening up a wider variety of spaces and opportunities. Indeed, locations close to areas of employment, economic opportunities, facilities and services are still significant for large groups of people. As Todes (2000) has observed, this debate has helped to avoid gross peripheralisation of the urban poor and brought questions of accessibility to the fore. Additionally, it has focused attention on the need to reconstruct township and informal areas and does, in part, weaken old divides. Pieterse (2003) has noted that perhaps the problem has less to do with integration as a strategy per se and that outcome failures need to be seen against the deeper problem of weak conceptual anchoring of the policy objectives and instruments, especially the failure to deal with divergent and conflictual interests in the city. We would also venture to add here that contrary to what the second wave of international research on urban policy in the developing world has had to say, that part of the problem may have to do with the fact that instead of semi-autonomously nurturing a desired form of the urban system that was robustly conceived in the early 1990s, policy makers have since subjugated the urban system to support the larger goal of accelerated economic growth which, whilst generating relative macroeconomic stability, has, to a large degree, failed to generate positive microeconomic developmental and employment advances. In other words, national urban policy goals may have been too embedded in overall development strategies and macro- and sectoral planning.

Ten years into democracy, the fact remains that a set of sustainable urban policies that talk to the multiple and competing demands and challenges of our cities has yet to emerge. Whilst positive inroads have been made, institutional fragmentation, competing discourses on development, limited policy and programme capacity at a local level, a lack of rigorous research in the preparation of policy and the temptation to invoke international best practice solutions in the reconstruction of our cities, has meant that we are at present, if not operating in a total policy void, reconfiguring our cities through ‘policy’ trajectories devoid of meaningful substance and content. It is a matter that needs urgent redress if the twin—and mutually reinforcing—concerns of growth and equity are to find expression in a newly formulated vision of the South African city.

An Overview—Post-1994

Introduction

There have been three components to urban policy in South Africa up to 2004. Policies to which close attention has been paid include those that gave effect to the ‘One city, one tax base’ slogan that emerged during the township struggles. These policies included re-demarcating municipalities to create integrated and democratic local governments, the comprehensive restructuring of the local government system and the design of municipal financial systems that support service delivery to the poor. Another set of policies revolves around the creation of ‘developmental local governments’ and includes integrated development planning and local economic development. A last set of policies refers to the mass delivery of free housing and services within municipalities.

National, provincial (in the case of housing) and municipal and sectoral policies over the past decade have sought to enable local government to undertake delivery, plan for delivery and implement delivery in consolidating democracy. Thus, government’s urban policy has focused on meeting the commitment in the Reconstruction and Development Programme (RDP) to provide for the basic needs of all South Africans and building democratic local government institutions and enhancing their ability to promote socioeconomic development in urban areas.

Backdrop

At the time of the 1994 democratic elections, South African cities were characterised by dire housing and services backlogs, inequalities in municipal expenditure, the spatial anomalies associated with the ‘apartheid city’, profound struggles against apartheid local government structures, high unemployment and many poverty-stricken households.

The African National Congress’ (ANC) commitment to addressing these issues can be traced to the 1994 RDP, which committed government to meeting the basic needs of all South Africans. Housing and services such as water and sanitation, land, jobs and others were counted as basic needs. The RDP also included the commitment to the restructuring of local government with a view to meeting these needs. The ANC recognised the key role of local government in delivering services and promoting economic development and called for the re-demarcation of local governments with a view to urban integration and democracy, the creation of a single tax base and the cross-subsidisation of municipal expenditure. Local governments were to become central to overcoming the backlogs.

However, the ANC confronted a fundamental difficulty. At the time, it was unknown how many households suffered from services backlogs, what household incomes were and what services levels they might afford, whether local governments had the capacity to deliver services as well as knowledge of alternative means of ensuring service delivery such as public–private partnerships, and how the capital and operating costs of the services might be financed. Indeed, there was only inexact data regarding the number of households in urban and rural areas. The same difficulties were not experienced in the case of housing, as the housing backlog had been estimated at the National Housing Forum (1992–1994) in the broadest possible terms as being 1.5 million units.

It was due to this lack of household and services information and a lack of clarity regarding options for delivering services that in 1995, the first version of the Municipal Infrastructure Investment Framework (MIIF) was prepared (Ministry in the Office of the President and the Department of Housing 1995). All too often based on informed guesstimates, the MIIF provided the ‘missing’ data and suggested how services might be delivered and financed.

At the same time as the MIIF was prepared, government was putting in place the preconditions for the ‘One city, one tax base’ policy: re-demarcating and creating integrated local governments in time for the 1995 and 2000 local government elections, with the latter being viewed as the ‘final stage’ in the creation of ‘developmental local government’. Examples of the policies and policy frameworks that emerged over time include integrated development planning, local economic development, free basic services and municipal services partnerships. Housing policy was an exception to this transformation because in 1992, negotiations began in the National Housing Forum, and by the time of the 1994 democratic elections, what was, in effect, a draft housing White Paper was in place.

Many urban policies have subsequently been revised in the light of experience and, importantly, also as government extended its democratic agenda. Free basic services provide an example. During the 1980s, one of the means employed to oppose the Black Local Authority system was a boycott on payment for rent and services. A decade and a half later, after it became clear that the widespread failure to pay for services showed no sign of stopping and also that many households could not afford to pay for services, the ANC and later government adopted the free basic services policy. This represented a break with the earlier principle included in the MIIF that taking into account the potential for cross-subsidisation, consumers of water and electricity should pay an amount for services consumed.

The Context for Urban Policy Formulation Post-1994

Population Growth and Urbanisation

In the first instance, the context for urban policy is the size of the urban population, its location, how rapidly it is growing and where it is growing. The South African census defines four types of areas: ‘tribal’ (former ‘homeland’), ‘rural formal’ or ‘commercial farming’, ‘urban formal’ and ‘urban informal’.¹

At the time of the 2001 census, South Africa’s population was 44,819,318, with about 57% of the population deemed to be urban and 43% rural. Forty-seven per cent of the urban population lived in formal urban areas and 8% lived in informal urban areas. Thirty-five per cent of the rural population lived in tribal areas and 7% in commercial farming areas. The 3% difference comprises overlapping urban and rural categorisation—institutional housing, hostels, industrial areas and smallholdings—with 2% being found in urban areas and 1% in rural areas.

The location of most of the urban population is depicted in Fig. 1, which shows the nine largest cities that are members of the South African Cities Network (SACN) and also provincial capitals that are not included in the SACN. The SACN cities are Johannesburg, eThekwini, Cape Town, Tshwane, Ekurhuleni, Nelson Mandela, Buffalo City, Msunduzi and Mangaung. In addition to the cities, Fig. 1 also reflects population density. Aside from Cape Town, it is apparent that most of the country’s population lives in the eastern half of the country.

Fig. 1

Cities comprising the SACN and provincial capitals. Source: HSRC

Figure 2 shows the population size and the number of households living in the cities included in Fig. 1. Johannesburg and eThekwini have more than three million inhabitants and Cape Town has close to that number. Ekurhuleni and Tshwane follow, with Ekurhuleni having about 2.5 million inhabitants and Tshwane two million inhabitants. In practice, the three cities in Gauteng comprise a single conurbation with a population of 7.7 million persons, approximating what the international literature has recently referred to as the emergence of global ‘city regions’ (Pillay 2004a, b). The last city with metropolitan status, Nelson Mandela, has about a million inhabitants, with Buffalo City, Mangaung and Msunduzi, not classified as metropoles, having much smaller populations. The smaller populations are also true of Polokwane and Mbombela, with a considerable drop to Mafikeng and Sol Plaatje. Sol Plaatje is located in the Northern Cape, which is losing population.

Fig. 2

Population and household numbers of SACN and selected secondary cities, 2001. Source: HSRC

In the case of the SACN cities, the most rapid growth is occurring in Gauteng. As can be seen in Fig. 3, the populations of Johannesburg, Ekurhuleni and Tshwane, at 22.2%, 22.4% and 18.0% between 1996 and 2001, respectively, are growing more rapidly than any other of the SACN cities (SACN 2004a, b: 38). At the same time, cities like Nelson Mandela, Msunduzi, Mangaung and Buffalo City are growing at a rate below that of the nation.

Fig. 3

Population and household growth rates of SACN cities, 1996–2001. Source: SACN (2004a, b: 38, 81)

The same data are unavailable for the other cities shown in Fig. 1 due to boundary changes. However, the populations of some other cities are also growing very rapidly, whilst others are experiencing a declining population. Examples of the former are Rustenburg in North West whose platinum mines are booming (4.9% per annum between the 1996 and 2001 censuses) and uMhlathuze on the KwaZulu-Natal east coast whose port and industrial activities are growing rapidly (8.07%). In contrast, in addition to Sol Plaatje (−0.28%), Mathjabeng (−3.07%) in the Free State is declining rapidly due to decline in the gold mining industry in the area (SACN 2004a, b: 38).

It is as a result of the different growth rates that data pertaining to cities and overall urbanisation should be read in conjunction with Fig. 4, which shows population growth between the two census periods, 1996 and 2001. Figure 4 points to the population decline in the Northern Cape and also many areas in the former homelands and also to the fact that in many areas, the rural population is growing more rapidly, often considerably more rapidly, than the national average of 2% per annum. Aside from growth in Cape Town, in high value mostly wine farming areas and along the coast north and west of the city, it is apparent that rapid growth is mostly occurring in the eastern and more northern parts of the country.

Fig. 4

Municipal population growth between 1996 and 2001. Source: HSRC

It appears that there is also rapid growth in some cities outside Gauteng and also in certain rural areas. The rural-to-rural migration is probably explained by the absence of jobs in the metros and larger urban centres, the availability of social grants that reduce household dependence on cities as possible sources of income and of subsidised housing and municipal and social services in small towns, predominantly in the commercial farming areas (Bekker and Cross 1999; Cross 2001).

A last observation regarding the location of urban growth is that most of the increase in the cities is occurring on the periphery of cities. This has long been known in the case of the provision since 1994 of subsidised low-income housing projects, which reinforce the disadvantages of poor access to jobs, social services and retail facilities, but this distant location phenomenon is also characteristic of new migrants. This situation is bemoaned by most urban practitioners and by the government itself in the Department of Housing’s 2004 Breaking New Ground policy statement where it is observed that:

After the 1994 elections, Government committed itself to developing more liveable, equitable and sustainable cities. Key elements of this framework included pursuing a more compact urban form, facilitating higher densities, mixed land use development, and integrating land use and public transport planning, so as to ensure more diverse and responsive environments whilst reducing travelling distances. Despite all these well-intended measures, the inequalities and inefficiencies of the apartheid space economy, has lingered on (2004: 11).

Household Growth and Housing and Services Backlogs

From the point of view of urban policy, the increase in the number of households is often more significant than the increase in the size of the population. This is because the increase in the number of households, together with household incomes, determines the housing backlog because service connections are made to houses, flats and so on, because household incomes determine the ability to pay for housing and services and because these have immediate implications for national budgeting for the housing subsidy and capital and operating subsidies, municipal finances and so on.

South Africa has experienced a sharp decline in household size and a consequent marked increase in the number of households. Between 1996 and 2001, the average number of households in the SACN cities grew by 27.5%, more than double the population growth rate. In 1996, the average household size was 4.47 persons; in 2001, it was four. If the household size had remained constant at the 1996 figure, the increase in the number of households would have been about 950,000. The actual increase was 2.13 million households, a difference of 1.18 million households.

Figure 3 shows population and household growth rates and Fig. 5 the differences in household size among the cities. Combining the two is important because rapid population growth coupled with declining household size accentuates the housing and services backlogs in an area. Again, it is the cities in Gauteng that stand out. The increase in the number of households in Johannesburg was a remarkable 8.05% per annum, about double the national average. Johannesburg, Ekurhuleni and Tshwane again have the most rapid growth rates. Indeed, with the exception of Buffalo City, the rate of increase in the number of households in the other SACN cities is below the national average. The rate of increase in the number of households in Nelson Mandela and Msunduzi in particular is sharply below the national average.

Fig. 5

Average household size of SACN and selected secondary cities, 2001. Source: HSRC

Possible explanations for the changes in household size include:

• Households that in the apartheid years suffered from massive overcrowding and combined, for example, married children with parents, are unbundling.

• The role of migration may also be a consideration. Cross (2001) suggests that a change in migration patterns is underway. Instead of circular migration, whilst many householders still talk of returning ‘home’ to rural areas, in practice, what is happening is that families are, de facto, separating into urban and rural households. This view is debated and to some degree supported by Russell (2002). However, Posel (2003) and Cox et al. (2004) disagree with the view that circular migration is giving way to separate urban and rural households.

• The impact of HIV/AIDS on households is unclear. AIDS deaths have picked up rapidly since 2001, but were not as profound between the 1996 and 2001 censuses (Dorrington et al. 2004: 24).

• South Africa has a very young population base, and the rate at which younger people enter the housing market exceeds the general populations’ growth rate, leading to smaller average household sizes.

• In regard to the influx of people from across South Africa’s northern borders, people might come in as individuals and not necessarily as families, thus reducing the average household size.

• Reportedly, the National Treasury has suggested that the availability of the housing subsidy has caused families to unbundle for the purpose of obtaining the subsidy.

The increase in the number of households among the SACN cities should be read in conjunction with Fig. 6 which points to the rapid growth of households, in excess of the national average of 4.34% per annum, in many areas of the country aside from the cities. Compared to areas where a population decline is shown (Fig. 1), there are fewer areas where there is a decline in the number of households. There is at the same time a slower than national average increase in the number of households in many of the former homelands.

Fig. 6

Household growth between 1996 and 2001. Source: HSRC

The implications of the increase in the number of households are evident in the housing backlog. Despite the delivery of more than 1.8 million subsidised housing units completed—or under construction—between 1994 and March 2005, in its 2004 Breaking New Ground document, the Department of Housing reported that the housing backlog had increased to more than 1.84 million units and is growing.

The proportions of households in the SACN cities that lack formal housing or water² on the stand or in the house are shown in Fig. 7. The proportion of households in the cities without formal housing ranges between 20% and 38%, with the three non-metropolitan cities and eThekwini and Ekurhuleni being at the top of the range. With the exception of Ekurhuleni, it is expected that this is partly because their demarcation included former homelands. The proportion of households in the cities without on-site water ranges between 15 per cent and 42 per cent, with eThekwini and the three non-metropolitan cities being at the top of the range. The data are a bit difficult to interpret. On the one hand, it makes sense that the cities whose boundaries include the former homelands will have the largest water backlogs; on the other hand, it is to be expected that formal houses will have on-site water, and this is not always the case, for example in Ekurhuleni. Rather than try to explain the latter differences, one wonders about the data.

Fig. 7

The percentage of households without formal shelter and on-site water in SACN cities, 2001. Source: SACN (2004a, b: 27–31, 80)

Population Projections and the Future Growth of the Cities

One cannot assume that high urban population and household growth rates will continue. South Africa’s population is growing increasingly slowly, although the projections that follow do not indicate that HIV/AIDS will lead to negative population growth. Increasingly, urban growth will be propelled less by the natural increase of the urban population and more by internal movement within the country and, perhaps, migration from other countries.

The population projections to 2015, including those for HIV/AIDS, are shown in Table 1 and are based on the Actuarial Society of South Africa (ASSA) 2002 model (Dorrington et al. 2004). The projections are for South Africa, and none are available for cities and for households. Due to the prevalence of HIV/AIDS, which the Nelson Mandela/Human Sciences Research Council (HSRC) survey reported as 11.4% in 2002, HIV/AIDS is a determining factor of population growth in South Africa.³

Table 1 Projected population, number of HIV positive, AIDS sick and cumulative AIDS deaths for 1990–2015, ASSA 2002 (default scenario)

The ASSA projections to 2015 require assumptions regarding:

• information and education campaigns;

• improved treatment of sexually transmitted diseases;

• voluntary counselling and testing;

• mother-to-child transmission prevention; and

• antiretroviral treatment (Dorrington et al. 2004: 8).

Regarding the overall expected trends for population size, Table 1 shows the total population for 5-year intervals, the annual population growth rate, the number of people infected with HIV and accumulated AIDs deaths. Dorrington et al. observe that:

The total population continues to increase over the period, although at a decreasing rate. From 2011, the expected annual rate of increase is 0.4%. The number of people infected with HIV peaks in 2013, at just over 5.4 million, after which it starts to decrease slowly. In contrast, the number of people sick with AIDS in the middle of each year continues to rise over the period, reaching nearly 743,000 in 2015. Accumulated AIDS deaths are close to 5.4 million by the same year. By 2004, it is estimated that over 1.2 million people have already died as a result of AIDS, just over 5 million are infected with HIV, and over 500,000 are AIDS sick (2004: 23).

According to these projections, population growth will slow from 0.7% per annum in 2005 to 0.3% in 2015. At the same time, because HIV infection most affects young adults, beginning at about 20 years old and continuing for another 20 or so years, due to the time lag between infection and death, it is in middle age that most deaths occur. This is when the dying are raising children and caring for the elderly. The death of one or more income-earning household members considerably increases the dependency burden on other household members. This, together with the large number of mostly working-age adults who are AIDS sick and in need of care, points to the burden that will be placed on households and the extended family.

Household Incomes

So far, the focus has been on the urbanisation of the population, the increase in the number of households, housing and services backlogs and the location of low-income households within cities. The significance of this material should be interpreted in the light of changing household incomes. This is because household incomes above a certain level enable the household to successfully participate in the private delivery of housing and services, contribute significantly to municipal rates and services, reduce the need for capital and operating subsidies for municipalities and so on.

In this regard, Table 2 provides extraordinary statistics regarding employment and unemployment amongst African and coloured people. Of the urban African and coloured labour force, 38.7% are without employment and 56.5% are without formal employment. Although there has been modest growth in formal sector employment, job creation in this sector has failed to keep pace with the growth in the labour force.

Table 2 Racial incidence of urban employment and unemployment, 2004

The negative implications for household incomes are ameliorated by the availability of social grants. In 1994, the government spent R10 billion on social grants, and there were 2.6 million beneficiaries. In 2003, the expenditure was R34.8 billion, and there were 6.8 million beneficiaries. It is speculated that it is due to these grants that there has been a decline in households with an income of less than R800 per month; conversely, it is speculated that it is due to increasing unemployment, that there has been an increase in households with an income of less than R3,500 per month.

Conclusion

The first 10 years of democracy have seen the creation of democratic, integrated and developmental local government, mass delivery of housing and services, a finely crafted array of capital and operating subsidies for delivery to low-income households and a number of programmes intended to enhance the capacity of local government to undertake delivery.

New Policies

The future agenda repeats the emphasis on delivery. For example, in A People’s Contract to create work and fight poverty, the ANC electoral platform for 2004, the ANC again commits government to the delivery of housing and services and then to reducing poverty and unemployment by half through speeding up the delivery of housing and services, ‘economic development, comprehensive social security, land reform and improved household and community assets…new jobs, skills development, assistance to small businesses, opportunities for self-employment and sustainable community livelihoods’.

There are four potentially significant new ‘urban policies’. These are the Presidency’s 2003 National Spatial Development Perspective (NSDP), the Department of Housing’s 2004 Breaking New Ground document, the Department of Provincial and Local Government’s Local Economic Development Framework and the pending, rewritten Urban Development Framework whose location within the Presidency, the Department of Provincial and Local Government or elsewhere had, at the time of writing, still to be determined. Each of these policies is briefly described and commented on later.

It does seem possible, though, that national policies of this sort may be of declining significance. The reason for this is due to the emergence of the SACN and the role of the major cities in setting their own and, in some respects, the country’s urban agenda. As a result, following the review of the four ‘policies’, the significance of the SACN and the metros is explored.

National Spatial Development Perspective

Within the NSDP, it is noted that it is not, in fact, a policy. Yet others, including prominent academics and consultants to government on urban policy, view the NSDP as an ‘incipient urbanisation policy’. Uncertainty is inevitable when the NSDP is presented as an:

…indicative guideline that will encourage creative interaction and coordination between departments and spheres of government about the nation’s spatial priorities. It will function as a basis for discussion and negotiation…the NSDP will function not as a policy that prescribes expenditure choices, but an instrument for discussing spatial development priorities for South Africa within government (Office of the President 2003: 38).

Quoting from the Atkinson and Marais chapter in this book:

The NSDP’s main argument is that areas with ‘potential’ or comparative advantage should be pinpointed, and thereafter receive priority in the allocation of resources—in particular, in the allocation of infrastructure funding (‘hard investments’). Government spending on fixed investment, beyond the obligation to provide basic services to all citizens, should therefore be focused on localities of economic growth and/or economic potential in order to attract private sector investment, stimulate sustainable economic activities and/or create long-term employment opportunities (Chapter 2).

There are similarities with the Department of Housing’s 1997 Urban Development Framework in respect of not being presented as a policy, not being prescriptive, but rather a ‘policy approach’ and a guide that will ‘foster linkages’.

A guide presented for discussion does not seem terribly promising unless there is a political obligation on relevant departments to take it forward. Encouragingly, President Thabo Mbeki mentioned the NSDP in his 2005 State of the Nation address. Reportedly, when national ministries are preparing their capital budgets and when provinces and municipalities are preparing their Growth and Development Strategies and their IDPs, questions are asked regarding whether these align with the NSDP. Eight provinces are now more explicit about spatial development perspectives in their provincial growth and development strategies. In effect, the intended and possible outcome of the NSDP may be horizontal coordination among national departments and vertical coordination between national, provincial and local spheres of government. Perhaps, the NSDP is beginning to serve as an urbanisation policy of sorts and will continue to do so for as long as there is political weight behind the guide.

But does the NSDP actually have significance in the case of housing and services? In the case of housing, the judgment of the Constitutional Court in the case of Government of the Republic of South Africa vs Grootboom 2001 SA 46 (CC)was that ‘the state’ must provide relief for people in ‘desperate need’ who are living in ‘intolerable conditions or crises’ situations. In recognition of the right to housing conferred in Section 26 of the Constitution, the Department of Housing is required to provide for ‘the fulfilment of immediate needs’. Similarly, the Constitution and the Grootboom judgment oblige all municipalities to provide services. It seems reasonable to assume that whereas the NSDP might influence the location of infrastructure investments for economic development purposes, it will have less bearing on the allocation and utilisation of resources in the scramble to provide serviced sites for housing and, in this respect, the realisation of A People’s Contract.

Breaking New Ground

The purpose of Breaking New Ground ‘is to outline a plan for the development of sustainable human settlements over the next 5 years, embracing A People’s Contract as the basis for delivery’ (Department of Housing 2004: 2).

Breaking New Ground is not presented as a new policy. Instead, it is noted that ‘Whilst Government believes that the fundamentals of the policy remain relevant and sound, a new plan is required…’ (Department of Housing 2004: 7). The document includes references to a ‘new vision’, ‘enhancement’, ‘amendments’, ‘changes’, ‘redirection’, ‘new systems’, ‘new policy measures’, a ‘new subsidy mechanism’ and a ‘new plan that will be required to redirect and enhance existing mechanisms’. President Thabo Mbeki’s view is that there should not be new policies or White Papers. Instead, the emphasis is on delivery, but Breaking New Ground arguably can be seen as a new housing policy.

The political underpinnings of Breaking New Ground separate this amended policy from the 1994 Housing White Paper where private sector agendas and technical experts held greater sway. Whereas a defining concern in housing policy was with the sustainability of the financial institutions, in Breaking New Ground, a defining concern is the sustainability of communities and settlements. The latter is in line with the Department of Housing being a signatory to the United Nations 1996 Habitat Agenda and the emphasis therein on sustainable settlements.

Breaking New Ground is a response to the sharp increase in the demand for housing arising from the decline in household size, increasing unemployment and numbers of households with incomes that qualify them for housing subsidies, the supply of housing on the urban periphery, with individual units not becoming “‘valuable assets’ in the hands of the poor” (Department of Housing 2004: 4) and a slow down in housing delivery and under-expenditure of provincial housing budgets.

Synthesising from Breaking New Ground, the government’s new housing vision requires accelerating the delivery of housing as a key strategy for poverty alleviation and job creation, creating assets, promoting social cohesion and improving quality of life for the poor, supporting the functioning of the entire single residential property market to reduce duality within the sector by breaking the barriers between the first economy residential property boom and the second economy slump, and building sustainable human settlements in support of spatial restructuring (emphasis in Breaking New Ground). This is a considerably more ambitious agenda than that in 1994 where the focus was simple: the delivery of a million units within 5 years to households having a monthly income below R3 500.

Local Economic Development Framework

At the time of writing, the Local Economic Development (LED) Framework was still unavailable. In the absence of the Framework document, the ensuing observations are based on the Department of Provincial and Local Government’s (2005: 1–3) LED implementation guidelines, which are to be included in the Framework document.

These guidelines were prepared following a consultative process involving provinces, (the South African Local Government Association) and key national sector departments through the Economic and Employment Cluster of Directors General. A series of consultative workshops were held with independent experts and in the international donor community. Discussions were also held with tertiary academic institutions and practitioners. A number of studies also within South Africa and abroad underpin these guidelines.

• 1. The aim of these guidelines is to influence the way government practitioners in all three spheres: understand, approach and implement LED in South Africa.

  The policy context for LED guidelines noted in the document combines the Constitution and the White Paper on Local Government with the mandate of the Department of Provincial and Local Government. The specific context for implementing LED consists of:

• 2. The Ten Year Review of government

• The 2004 electoral mandate

• The Medium Term Strategic Framework

• The January 2005 Cabinet Lekgotla

• The Microeconomic Reform Strategy

Further:

• 3. A government-wide approach to developing and supporting robust and inclusive municipal economies is required and should be facilitated through the active and dynamic alignment of the NSDP, Provincial Growth and Development Strategies (PGDSs), and District/Metro Integrated Development Plans (IDPs; and their LED programmes).

If one considers this rather demanding context for local economic planning and the desire for horizontal coordination among national government departments, as well as vertical coordination within the three spheres of government, one can understand why there is so much attention in these various ‘policy’ documents to ‘alignment’, ‘coordination’, ‘harmonisation’, ‘integration’ and so on. This desire for coordination may become problematical when one notes that economic development in the cities relies on private investment, which may be little influenced by national policy for LED.

Urban Development Framework

The 2005 draft Urban Development Framework appears to be an attempt to resurrect the 1997 Urban Development Framework. However, it is unclear whether the Framework will actually be finalised and agreed to by Cabinet. This brief assessment is based on an earlier draft of what was then called an Urban Policy/Strategy. The first observation is, of course, that the significance of the document is diminished by its having been downgraded from a policy, which requires the approval of Cabinet, to a Framework that requires the approval of the minister of the department concerned.

The purpose of the draft Urban Policy/Strategy is to ‘promote’, ‘initiate’ and ‘propose’ (1) a perspective: ‘initial set of practical interventions for investigation’, ‘deeper understanding’, ‘focus on certain policy issues’, ‘debate’, ‘developing a shared vision’ and (2) coordination: ‘integration and improved service delivery’, ‘alignment of government policies and programmes’, ‘appropriate funding framework’.

The intention, again, is to promote coordination, integration and alignment with national political commitments such as A People’s Contract.

The Agenda of the Cities

The counterpoint to the national perspective and frameworks is the agenda of the cities, especially those comprising the SACN. The SACN and its members have an increasingly robust role in setting the urban agenda and, in effect, leading urban policy. Many points contribute to this view.

First, the context for urban policy is ambivalence within government regarding the preparation of new urban policies. Reportedly, it was following debate within Cabinet that the NSDP was termed a perspective and not a policy. Breaking New Ground is presented as an ‘amendment’ or ‘enhancement’ to existing policy. The LED policy was downgraded to a framework. The Urban Development Framework is, of course, a framework. This national policy-shy approach to urban policy is in part a reaction to past accusations of government policy favouring urban areas (Development Bank of Southern Africa 2005).

In contrast, in its State of the Cities Report, the SACN, “Through description and analysis of trends…hopes to set up a strategic agenda for further research, planning and action…” (SACN 2004a: 5). In addition, with the assistance of the World Banklinked Cities Alliance, in its 2004 People and Places: An Overview of Urban Renewal, the SACN unabashedly specifies that, “A national urban renewal policy framework must be developed as part of a broader South African urban policy framework” (SACN 2004b: 9), and, in the absence of a national policy framework, the document also indicates areas where cities can themselves take aspects of the policy agenda forward. It is further indicated that this should take the form of a City Development Strategy. City Development Strategies are explained as:

A city development strategy supports cities in this critical decision-making process and is focused on implementation. It is an action-plan for equitable growth in cities and their surrounding regions, developed and sustained through participation, to improve the quality of life for all citizens. The output of a city development strategy includes a collective city vision and a strategic action plan aimed at policy and institutional reforms, increased economic growth and employment, and implementation and accountability mechanisms to ensure systematic and sustained reductions in urban poverty⁴ (emphasis added).

As prescribed in City Development Strategies, the cities are taking their development forward in partnership with the private sector. The significance of the relationship with the private sector is that it further emphasises the independence of the cities from alignment with the national policy direction.

The presumptions and prerogatives of the SACN cities are further contributed to by the ‘normalisation’ of the urban agenda, with the legislative and institutional prerequisites for dealing with the aftermath of urban apartheid having been put in place. For example, the SACN now turns its attention to developing instruments of urban governance such as ‘an effective regulatory system for land use planning that addresses the realities of informal settlements’ (Boraine et al. 2005: 4).

To this should be added the role of municipalities in the delivery of housing and services. The Constitution, regardless of national urban policy, amendments and frameworks, requires that municipalities invest in services infrastructure for delivery to low-income households. This requirement is abetted by the Municipal Infrastructure Grant, which came into effect in 2004–2005, which accords municipalities increasing independence in the allocation of resources for investment in services infrastructure. The same can be said for the Department of Housing’s intention to accredit the cities to deliver houses.

Then, to add to the urban voice, there is increasing recognition of the facts that poverty is not solely a rural issue and there is an equivalent, if not greater, prevalence of poverty in urban areas and that the cities play the central role of driving economic growth and employment creation (Development Bank of Southern Africa 2005: Chapter 5). Well-run cities are a precondition for both competitiveness in the global economy and for alleviating poverty in both urban and rural (owing to remittances and migration) areas.

The point is that in a context of ambivalence at the national level in relation to urban policy, the cities comprising the SACN are proceeding with an urban agenda that, to a significant degree, is self-defined, enabled by national housing and services policies and subsidy frameworks and embodied in their commitment to city development strategies. The point is also that the SACN agenda influences that of non-member secondary cities. A useful window through which to examine this development is in the planning and preparation ahead of the FIFA World Cup in 2010. Whilst a national planning blueprint exists (derived from the 2004 Bid blueprint) and a number of national government departments have assumed responsibility for driving facets of the planning, it is really the 10 hosts cities themselves, liaising closely with the Local Organising Committee and FIFA, that are driving this process. Given strict timeframes and an unforgiving set of obligations and specifications from FIFA, it seems prudent that the nine host cities are advancing the planning process in earnest. In doing so, they seem to be further consolidating a trend where cities themselves have become the lead agents in most facets of urban development, regeneration and renewal.